This script benchmarks CPU performance on OpenWrt devices using stress-ng.
It measures Bogo Ops/s for int64 and matrixprod methods on both single-thread and multi-thread execution.
The results are printed as a Markdown-formatted table row, ready to be copied into the table below.
Feel free to submit a pull request with benchmark results from your device.
Note
Tested only OpenWrt
sh <(wget -O - https://raw.githubusercontent.com/itdoginfo/cpu-wrt-bench/refs/heads/main/cpu-wrt-bench.sh)
| Device | CPU | OpenWrt Version | CPU Threads | int64 (1 Thread) | matrixprod (1 Thread) | int64 (ALL Threads) | matrixprod (ALL Threads) |
|---|---|---|---|---|---|---|---|
| Beeline SmartBox TURBO+ | MediaTek MT7621A | OpenWrt 24.10.1 | 4 | 35.77 | 7.88 | 92.83 | 17.94 |
| Xiaomi Mi Router AX3000T | MediaTek MT7981BA | OpenWrt 24.10.1 | 2 | 291.67 | 10.89 | 581.33 | 19.86 |
| FriendlyElec NanoPi R2S | Rockchip RK3328 | OpenWrt 24.10.1 | 4 | 286.60 | 10.80 | 1144.74 | 35.30 |
| Xiaomi Redmi Router AX6000 | MediaTek MT7986A | OpenWrt 24.10.1 | 4 | 451.73 | 18.78 | 1805.96 | 55.30 |
Note
n/s - CPU not supported int128 method.